Alternanthera philoxeroides is a herbaceous perennial plant, under normal conditions, it can spread quickly via clonal growth (Luo et al., 2009; Ayi et al., 2016), growing to a height of 50–120 cm, with a long single or sparsely branched stem. The stem has several internodes of different lengths, and its maturity degree gradually decreases from the base to the top, making it an ideal organ type to study the adaptation and response of A. philoxeroides to different submergence depths.

To reduce the influence of environmental conditions on mother plants, A. philoxeroides plants used in this experiment were cultivated from cuttings obtained from plants naturally growing on the banks of the Jialing River in Chongqing, Southwest China (29°49′N, 106°25′E). In May 2020, unbranched plants with a stem length of ~30 cm were selected and cut at the stem base. The cuttings were transported to the laboratory immediately, and healthy and vigorous cuttings were selected for subsequent treatments, all trimmed to have five internodes and six leaves. Each selected cutting was planted in a plastic pot (diameter and depth were both 13 cm) containing riparian soil from the Jialing River banks, two stem nodes of the cutting (hereafter referred to as “plant”) were buried in soil for rooting. All plants were cultivated under the same conditions and placed in an open field of the experimental garden affiliated to the Key Laboratory of Eco-environments in Three Gorges Reservoir Region (Ministry of Education) at Southwest University, Chongqing. The temperature, relative humidity, daily maximum light (PAR) intensity, and water provision were maintained at 10–15°C, 75–85%, 600–800 μmol m⁻² s⁻¹, and ~80–90% of soil water-holding capacity, respectively. Plants were watered daily. Lateral buds, if produced, were removed to ensure that the main stem of plants grew without branching. Plants were kept growing upright by the support of thin bamboo sticks. After ~1 month of cultivation, plants with ~45 cm in height and 12 internodes were selected for submergence treatments.

Four submergence treatments were applied in a fully randomized design using selected plants (20 replicates per treatment). Submergence depth of 0 m was set as control: unsubmerged plants were placed under dark conditions and watered normally to keep soil water at field capacity. Additionally, three groups of plants were submerged in a water-filled concrete reservoir (length × width × depth = 2.3 × 1.9 × 10 m), with the top of plants 2, 5, and 9 m beneath the water surface. The plants in pots were suspended at planned water depths (Figure 1A). Pilot experiments showed that the stem tips of plants started to die (characterized by becoming flaccid) on the 7th day following submergence at depth of 9 m. Therefore, the treatment duration was set at six days for all submergence depths to ensure tested plants kept vigorous during this period.

To investigate the effects of submergence depth on plants, the physico-chemical status of water (light, dissolved oxygen (DO), pH, and temperature) in the concrete reservoir were kept constant at any depths. The water was air-saturated by pumping air every day through an air pipe with vent holes installed at the bottom of the reservoir, ensuring the adequate and uniform supply of oxygen and carbon dioxide at various water depths. The temperature of the whole water body was kept constant by applying an electric heating system equipped at the bottom of the reservoir. The reservoir was covered with black sun-shading nets to eliminate the influence of light. The control treatment was conducted in darkness under the same temperature as the submergence treatments. DO concentration, photosynthetically active radiation (PAR) intensity, temperature, and pH of the water column at different depths in the reservoir were checked twice per day (morning and evening) using a multi-parameter water quality analyzer (Hydrolab DS5, Hach, United States). During the experiment, no significant difference in these factors was found among different water depths (Table 1).

Each plant had ~12 stem internodes at the start of treatments. From the stem base upwards, the 1st to 6th internodes were mature and the 7th to 12th internodes were immature (Figure 1B). We marked non-destructively immature internodes so as to distinguish between the mature, immature internodes formed before treatment and the new internodes produced during treatment. At the beginning of treatments, for each plant, 26 leaves were retained on the upper stem by trimming other leaves from the stem base upwards. The length of all internodes of each plant was measured twice (at the beginning and the end of treatments). We compared the elongation of mature, immature, and newly produced internodes at the end of the treatments to assess the effects of submergence depths on the growth of internodes of varying maturity degrees. The elongation of mature internodes, immature internodes, and newly produced internodes was, respectively, calculated as the sum of elongation from the 1st to 6th internodes, the sum of elongation from the 7th to 12th internodes, and the length sum of all newly produced internodes (Figure 1C). The number of newly produced internodes was also counted. Afterward, the leaves, stem (viz. the aboveground stem part), and underground part (including the underground stem part which can be regarded as rhizome, and roots) of each plant were oven-dried to constant weights at 75°C, and their dry mass was determined (BSA124S, Sartorius, Germany).

No adventitious roots were formed on the aboveground stem nodes of A. philoxeroides before the submergence treatments. Any formation of aquatic adventitious roots on the aboveground stem nodes was recorded at the end of the treatments.

Because the pilot experiment showed that the 10th stem internode had relatively large elongation during treatments (Supplementary Figure S1), and its pith cavity had not been formed at the start of the treatments, we marked the 10th internode before treatments to investigate the effects of submergence depths on the development of internodal pith cavity. Transverse sections were made at the middle of the internodes immediately when treatments terminated (Figures 1D,E). Sections were observed and photographed using a stereomicroscope (SMZ25, Nikon, Japan). The pith cavity cross-area of each section was measured using the NIS-elements imaging software (version 4.30).

Soluble sugars and starch of the dried stems and underground parts of plants were measured by using fine powder samples prepared with a ball mill (WS-MM200, Retsch, Haan, Germany). Soluble sugars and starch were extracted and determined using a modified method based on the traditional anthrone-sulfuric acid method (Zhang, 2003; Lei et al., 2014). Ethanol-soluble sugars of 0.01 g powder sample soaked in 80% (v/v) ethanol solution were extracted in the water bath at 80°C for 40 min. The extraction of each sample was repeated twice and the extracts were mixed and diluted with ultrapure water to 50 ml. Subsequently, the residue of each sample was soaked in 5 ml ultrapure water and extracted for soluble but ethanol-insoluble sugars in the water bath at 80°C for 40 min, the extraction of each sample was also repeated twice and the extracts were mixed and diluted with ultrapure water to 50 ml. The starch of the residue was hydrolyzed to soluble sugars using 6 mol L⁻¹ HCl, the hydrolyte solution was filtered into a flask and diluted with ultrapure water to 100 ml. A total of 1 ml of the above-mentioned 50, 50, and 100 ml diluted extracts was respectively added with 5 ml anthrone-sulfuric acid reagent (0.1 g anthrone dissolved in 100 ml 75% (v/v) sulfuric acid solution) and heated for 10 min at 100°C, and the light absorbance at 625 nm was measured (UV-2700, Shimadzu, Japan) to determine the concentration of ethanol-soluble sugars, soluble but ethanol-insoluble sugars, and starch based on a glucose calibration curve. The concentration of total soluble sugars in a plant sample was the concentration sum of ethanol-soluble sugars plus soluble but ethanol-insoluble sugars. Since the starch was hydrolyzed to soluble sugars, starch concentration was therefore calculated by multiplying the concentration of hydrolytic soluble sugars with a hydrolysis coefficient (0.9). The sum of ethanol-soluble sugars, soluble but ethanol-insoluble sugars, and starch was regarded as non-structural carbohydrates in plant samples.

The effects of treatments (control and submergence at different depths) on stem and internode elongation, plant and organ biomass, number and length of newly produced internodes, internodal pith cavity cross area, number of nodes forming adventitious roots, and contents of non-structural carbohydrates (soluble sugars and starch) were checked using one-way ANOVA. Logarithm transformation of data was performed to equalize variance if necessary. Differences between treatments were detected using Duncan's multiple range test, and the significance level was set at p = 0.05. All the analyses were conducted using SPSS 22 (SPSS Inc., Chicago).

A. philoxeroides plants subjected to four treatments [water depths of 0 m (control), 2, 5, and 9 m] all elongated their stems during the experiment; however, stem elongation significantly decreased with increasing submergence depth after 6 days of treatment (Figure 2A). The stem elongation of plants subjected to submergence at a water depth of 0 (control), 2, 5, and 9 was 19.79, 28.56, 23.21, and 11.99 cm, respectively. Plants submerged at water depths of 2 and 5 m presented significantly larger stem elongation than control plants, but plants submerged at water depth of 9 m presented much less stem elongation than control plants (Figure 2A). The contribution of mature and immature internodes produced before treatments and new internodes produced during treatment to the stem elongation differed significantly, immature internodes comparatively made the largest contribution to plant stem elongation (Figure 2B).

The elongation responses to submergence depths differed between mature, immature, and newly produced internodes. The total elongation of mature internodes was not affected by submergence depths, but the total elongation of immature internodes and newly produced internodes differed significantly between submergence depths (Figure 2B). The total elongation of immature internodes were 14.86, 20.75, 16.95, and 9.48 cm in plants submerged at water depths of 0 (control), 2, 5, and 9 m, respectively (Figure 2B).

Very few new internodes were produced during submergence at water depth of 9 m, and 1.85, 2.40, and 1.85 new internodes on average were produced at water depths of 0, 2, and 5 m, respectively (Table 2). The total length of new internodes was significantly different between the treatments, with the largest length realized in plants submerged at depth of 2 m (Figure 2B).

Based on the results of this study, it was found that the stem elongation and internode production of submerged A. philoxeroides decreased with increasing submergence depths, with the largest stem elongation and new internode production showing under the shallowest submergence (2 m depth) and the smallest stem elongation and nearly no new internode production under the deepest submergence (9 m depth). However, in comparison with plants unsubmerged, the stem elongation and internode production of plants were enhanced by submergence at depths of 2 and 5 m.

The internodal pith cavity developed during the experiment varied among four treatments (Figure 3). Plants submerged at depths of 2 and 9 m had the broadest and the narrowest internodal pith cavity, respectively. Plants submerged at depth of 5 m did not differ from unsubmerged plants in size of internodal pith cavity (Figure 3).

The number of stem nodes forming adventitious roots upon submergence decreased significantly with increasing submergence depth (Figure 4). Approximately 4–9 nodes (an average of 6.5) on plants submerged at depth of 2 m formed adventitious roots, and 0–5 nodes (an average of 1) on plants at a submergence depth of 5 m formed adventitious roots, only 1 node of 1 plant (20 plants in total) formed adventitious roots when submerged at depth of 9 m (Figure 4).

It was obvious that increasing water depth impeded the pith cavity development and adventitious root formation of submerged A. philoxeroides plants.

At the end of the experiment, the total biomass of submerged A. philoxeroides plants decreased gradually with increasing submergence depth, however, unsubmerged plants and plants submerged at a depth 2 m did not differ in total biomass (Figure 5A). The total leaf mass of plants submerged at a depth 9 m was the smallest among all treatments, but the total leaf mass did not differ between plants submerged at depths of 2 and 5 m, and plants unsubmerged. The largest and the smallest stem mass presented respectively in plants submerged at depths of 2 and 9 m, and plants submerged at depth of 5 m had similar stem mass to control plants (Figure 5B). Unlike the responses of stem and total leaf mass to treatments, the underground mass of plants declined with increasing submergence depths, plants unsubmerged tended to achieve the largest underground mass, and plants submerged at a depth 9 m had the smallest underground mass (Figure 5B).

The non-structural carbohydrates of A. philoxeroides plants were mainly composed of soluble sugars. At the end of the experiment, the concentration of soluble sugars in both stems and underground parts of submerged plants decreased with increasing submergence depth (Figures 6A,B). As to the concentration of soluble sugars in stems, unsubmerged plants were not different from plants submerged at depth of 5 m but were lower than plants submerged at depth of 2 m and higher than plants submerged at depth of 9 m (Figure 6A). Unsubmerged plants were similar to plants submerged at depth of 2 m in the concentration of soluble sugars in the underground plant parts but had a higher concentration of soluble sugars than plants submerged at depth of 5 and 9 m (Figure 6B).

No significant difference was found between treatments in starch concentration in stems of A. philoxeroides (Figure 6C). But starch concentration in underground plant parts tended to increase with increasing submergence depth, plants submerged at depth of 9 m had the highest concentration of starch (Figure 6D).

A. philoxeroides exhibited a typical “escape strategy” when submerged at depth of 2 m via fast shoot elongation by producing new internodes and elongating its immature internodes (Figure 2, Supplementary Figure S1, Table 2). Although previous studies showed that escape strategy was mainly presented in plants submerged partially or in waters shallower than 2 m depth (Bailey-Serres and Voesenek, 2008; Manzur et al., 2009; Akman et al., 2012), our present work indicated that escape strategy can also be selected by A. philoxeroides under submergence depths equal or larger than 2 m, which implies that A. philoxeroides can escape from the complete submergence with a water depth of 2 m or even deeper. Theoretically, the fast elongation of stems under submergence at depth of 2 m might be lethal to plants because of the substantial energy consumption and consequent energy crisis. However, our experimental results showed that the biomass and carbohydrate consumption of plants at 2 m submergence depth did not differ from those of unsubmerged plants at the end of the experiment (Figures 5, 6). This might be due to the morpho-physiological behaviors the plant took to help mitigate possible damages, especially energy crisis. First, adventitious root formation was enhanced in A. philoxeroides plants submerged at depth of 2 m (Figure 4), which was beneficial for plants to improve the supply of oxygen and mineral nutrients under submergence stress, because adventitious roots are able to absorb oxygen and mineral nutrients from water (Ayi et al., 2016, 2019). Second, the plants submerged at depth of 2 m developed the widest internodal pith cavity (Figure 3), which surely facilitated plants' internal gas transport (Jackson and Armstrong, 1999; Pedersen et al., 2021). Therefore, our results indicated that A. philoxeroides was able to resist submergence of 2 m depth by adopting an escape strategy via stem elongation, pith cavity development, and adventitious roots formation.

When submerged at depth of 9 m, A. philoxeroides adopted a quiescence strategy by minimizing its growth, including almost no new internode production (Table 2) and new internode elongation (Figure 2), suppressed pith cavity development (Figure 3) and adventitious root formation (Figure 4). This quiescence had a “two-edged sword” effect. On one hand, the weakened growth of new tissues and organs led to low consumption of substances including carbohydrates; on the other hand, suppressed pith cavity development and adventitious roots formation decreased the utilization efficiency of carbohydrates under deep submergence conditions. The former would reduce the carbohydrate requirement of submerged plants and enhance their submergence tolerance; the latter may lead to fast carbohydrate consumption, which would be lethal under prolonged submergence. Our results showed that the loss of total biomass in plants submerged at depth of 9 m was significantly higher than that in unsubmerged plants (Figure 5), and the concentration of soluble sugars was lower in plants submerged at depth of 9 m than in unsubmerged plants (Figures 6A,B). Nevertheless, all plants submerged at depth of 9 m survived with no tissue corruption, indicating the high submergence tolerance of A. philoxeroides.

The responses of A. philoxeroides to submergence at depth of 5 m indicated that the plant was probably in the process of strategy shift from escape to quiescence. Plants submerged at depth of 5 m presented larger stem elongation than unsubmerged controls (Figure 2), which was an obvious sign of escape strategy. However, the reduced adventitious root formation, decreased total plant biomass, and lowered soluble sugars concentration in the underground parts of plants submerged at depth of 5 m (Figures 4–6), as compared to those of unsubmerged controls, revealed quiescence strategy was also adopted by plants submerged at depth of 5 m. Thus, we inferred that A. philoxeroides may gradually switch its strategy from “escape” to “quiescence” as a response to submergence with increasing depth. It was reported that Lotus tenuis could quickly switch from an escape strategy under partial submergence to a quiescence strategy when confronted with complete shallow submergence (Manzur et al., 2009). Obviously, A. philoxeroides did not take this quick strategy switchover but performed gradual change in strategy when submergence depth altered.

Previous studies have pointed out that A. philoxeroides is a representative species that adopts an escape strategy in response to submergence (Luo et al., 2009, 2011; Ayi et al., 2016). This study supported this argument. However, it was also found in this study that A. philoxeroides was able to change its strategy from “escape” to “quiescence” under deep submergence (deeper than 5 m in this study), an intriguing phenomenon that was not observed in previous studies. Based on our experimental results, two questions need to be more focused on in the following studies are raised: (1) do all species that typically adopt escape strategy automatically alter their strategy with increasing submergence depth and (2) what role does the submergence depth play in the strategy switchover?

Low light, whether in terrestrial or in aquatic habitats, is an inducing factor for plants to adopt an escape strategy (via fast shoot and petiole elongation) due to shade avoidance (Mommer et al., 2005; Pierik et al., 2011; Sasidharan et al., 2014). In this study, the stem elongation of A. philoxeroides submerged at depths of 0, 2, and 5 m in darkness substantiated this statement (Table 2; Figure 2); however, the retarded growth of A. philoxeroides submerged at depth of 9 m in the same darkness (Table 2; Figure 2) suggested that the light availability can not sufficiently explain the strategy change along a gradient of submergence depths. If the stem elongation upon submergence were induced by low light or darkness due to phototaxis of plants, plants submerged at any depths in darkness should present similar stem elongation.

Generally, the escape and quiescence strategies represent the syndrome that plants exhibit as a response to low oxygen levels in water (Voesenek and Bailey-Serres, 2013, 2015). Oxygen shortage might induce the production of reactive oxygen species (Paradiso et al., 2016; Sasidharan et al., 2018), nitric oxide (Mugnai et al., 2012; Paradiso et al., 2016), ethylene (Fukao and Bailey-Serres, 2008; Sasidharan et al., 2018), and other signaling molecules (Sasidharan et al., 2018), as well as activate a series of phytohormone-synthesizing molecules (Cox et al., 2004) to induce different strategies among species. In plant species adopting an escape strategy, the low oxygen level not only stimulates the elongation of shoots and petioles, but also induces aerenchyma formation, via increasing porosity in the roots or stem (Parlanti et al., 2011; Pedersen et al., 2021), widening the pith cavity (Steffens et al., 2011), and forming adventitious roots (Zhang et al., 2015b; Ayi et al., 2016). However, in a water body with dissolved oxygen saturated (like the water environment A. philoxeroides experienced in this study), it is hard to attribute the strategy shift of plants submerged at different water depths to oxygen availability.

In our experimental system, all factors except water depth were kept constant (Table 1). Therefore, it is logical to infer that water depth was linked to the strategy changeover of A. philoxeroides. It was found in previous studies that water depth affected plant distribution via hydrostatic pressure (Dale, 1981; Makarov, 2011). Some studies suggested that low hydrostatic pressure could accelerate physiological activities and stimulate plant growth, whereas high hydrostatic pressure may affect the stabilization of cells or enzymes, depress physiological activities, and impede cell division or elongation (Adkins et al., 1990; Vervuren et al., 2003; Yi et al., 2016; Bejarano et al., 2018). The strategy changeover of A. philoxeroides along submergence depth gradient in this study was very probably caused by the change in hydrostatic pressure, but how hydrostatic pressure induces strategy changeover needs to be clarified.